Sardana: an Automatic Tool for Numerical Accuracy Optimization
Abstract
On computers real numbers are approximated by floating-point numbers defined by the IEEE754 formats [1]. For most computations these formats are precise enough even though the induced approximation errors inherently. However in some cases the accuracy of the calculation is critical and the user needs to certify that his program will always yield an accurate enough output for every valid input. As it is impossible to check the validity of a calculation for every inputs, static analyzers such as Fluctuat or Astr ee rely on an interval or relational representation of the inputs, combined to abstract interpretation.
Domains
Computer Arithmetic
Origin : Files produced by the author(s)
Loading...